Key Laboratory of Ministry of Education for Image Processing and Intelligent Control

HUST, China

Key Laboratory of Ministry of Education for Image Processing and Intelligent Control

HUST, China
SEARCH FILTERS
Time filter
Source Type

Li M.,Huazhong University of Science and Technology | Li M.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Yang Y.-P.,Huazhong University of Science and Technology | Yang Y.-P.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | And 2 more authors.
Zidonghua Xuebao/Acta Automatica Sinica | Year: 2010

Graph-based semi-supervised methods define a graph where the nodes are labeled and unlabeled examples in the dataset, and edges reflect the similarity of examples. These methods usually assume label smoothness over the graph. Graph methods are nonparametric, discriminative, and transductive in nature. These methods take high classification accuracy on variant data distributions. But the computation complexity is very high. As the size of dataset grows, the graph will be too large to compute and this limits the extension of its usage. In this paper, we propose a novel method for fast computation based on local clustering, which is very efficient for reduction of graph size and can maintain the accuracy at the same time. The local clustering method is of low computation complexity and the data structure can be preserved by a newly designed distance function. Experimental results show that this approach can preserve the accuracy of purely graph-based methods and significantly reduce computational cost. Copyright © 2010 Acta Automatica Sinica. All rights reserved.


Gao J.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Huang X.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Liu B.,BFSU
Machine Vision and Applications | Year: 2010

To improve the real-time performance, a quick scale-invariant interest point detecting approach based on the image color information is proposed in this paper. The approach uses the scale normalized Laplacian operator to extract the interest points in the incomplete image pyramid. A new local descriptor is presented in the approach to compute the feature vector of each interest point. The descriptor is made up with several subregions like the SIFT (Scale-Invariant Feature Transform) descriptor, meanwhile, it chooses the mean values of different color components in each subregion as the feature vector's elements to differentiate color objects better and reduce the descriptor's dimension. Through the experiment, the detected interest points are robust to many image transformations and the approach is indicative of needing less computation than other interest point detecting algorithms. The research discloses that the approach can obtain both superior stability and real-time performance at the same time. © Springer-Verlag 2008.


Huang S.,Huazhong University of Science and Technology | Huang S.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Zhou C.,Huazhong University of Science and Technology | Zhou C.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | And 7 more authors.
Reliability Engineering and System Safety | Year: 2016

Brake-by-wire (BBW) systems that have no mechanical linkage between the brake pedal and the brake mechanism are expected to improve vehicle safety through better braking capability. However, transient faults in BBW systems can cause dangerous driving situations. Most existing research in this area focuses on the brake control mechanism, but very few studies try to solve the problem associated with transient fault propagation and evolution in the brake control system hierarchy. In this paper, a hierarchical transient fault tolerant scheme with embedded intelligence and resilient coordination for BBW system is proposed based on the analysis of transient fault propagation characteristics. In this scheme, most transient faults are tackled rapidly by a signature-based detection method at the node level, and the remaining transient faults, which cannot be detected directly at the node level and could degrade the system performance through fault propagation and evolution, are detected and recovered through function and structure models at the system level. To jointly accommodate these BBW transient faults at the system level, a sliding mode control algorithm and a task reallocation strategy are designed. A simulation platform based on Architecture Analysis and Design Language (AADL) is established to evaluate the task reallocation strategy, and a hardware-in-the-loop simulation is carried out to validate the proposed scheme systematically. Experimental results show the effectiveness of this new approach to BBW systems. © 2016 Elsevier Ltd. All rights reserved.


Shen Y.,Huazhong University of Science and Technology | Shen Y.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Zhenzhen J.,Huazhong University of Science and Technology | Zhenzhen J.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | And 4 more authors.
Chinese Control Conference, CCC | Year: 2015

Two types of novel cellular neural networks based on mem-elements are proposed, namely, MC-CNN and EM-CNN. The MC-CNN lets a memcapacitor replace the conventional linear capacitor of a cellular neural network cell. This improvement takes advantage of the nanoscale of memcapacitor and its natural nonlinearity, which makes the MC-CNN more compact and the output function simplified. Mathematical analysis of stability and simulation of image processing is presented to verify the feasibility and performance of MC-CNN. The EM-CNN is an economical improvement of the memristor synapse cellular neural network. In the EM-CNN, based on the symmetry of CNN templates, the amount of memristors and voltage-controlled current source is largely reduced. Thus, the EM-CNN is not only economical on the fabricating cost of CNN but also possesses a simpler cell structure which is beneficial to better implementation of CNN. © 2015 Technical Committee on Control Theory, Chinese Association of Automation.


Hu X.,Huazhong University of Science and Technology | Hu X.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Xiong W.,Huazhong University of Science and Technology | Li W.,Huazhong University of Science and Technology | Ke L.,Huazhong University of Science and Technology
Chinese Control Conference, CCC | Year: 2015

The rapid progress in the research and development of electronics, sensing, signal processing, and communication networks has significantly advanced the state of applications of intelligent transportation systems (ITS). However, efficient and low-cost methods for gathering information in large-scale roads are lacking. Consequently, wireless sensor network (WSN) technologies that are low cost, low power, and self-configuring are a key function in ITS. The current state of the art of application scenarios of WSNs for urban transportation is captured in this paper, where solutions are discussed under their related application schemes. This paper also points out the potential application scenarios and design requirements of WSN for urban transportation. © 2015 Technical Committee on Control Theory, Chinese Association of Automation.


Wang G.,Huazhong University of Science and Technology | Wang G.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Yin Q.,Huazhong University of Science and Technology | Yin Q.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | And 3 more authors.
Circuits, Systems, and Signal Processing | Year: 2013

This paper is concerned with the H ∞ synchronization problem of directed complex dynamical networks. For the system under study, mixed time-delays and stochastic factors are taken into consideration. By stochastic analysis techniques, we establish sufficient conditions under which the network is synchronized to the prescribed system in the H ∞ sense. Moreover, the effect of exogenous disturbance on synchronization performance is reduced to a given H ∞ norm. Finally, numerical simulations are given to verify the usefulness and effectiveness of our results. © 2012 Springer Science+Business Media New York.


Wang G.,Huazhong University of Science and Technology | Wang G.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Shen Y.,Huazhong University of Science and Technology | Shen Y.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | And 2 more authors.
Chinese Physics B | Year: 2013

This paper is concerned with the exponential synchronization problem of coupled memristive neural networks. In contrast to general neural networks, memristive neural networks exhibit state-dependent switching behaviors due to the physical properties of memristors. Under a mild topology condition, it is proved that a small fraction of controlled subsystems can efficiently synchronize the coupled systems. The pinned subsystems are identified via a search algorithm. Moreover, the information exchange network needs not to be undirected or strongly connected. Finally, two numerical simulations are performed to verify the usefulness and effectiveness of our results. © 2013 Chinese Physical Society and IOP Publishing Ltd.


Huang S.,Huazhong University of Science and Technology | Huang S.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Zhou C.-J.,Huazhong University of Science and Technology | Zhou C.-J.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | And 3 more authors.
International Journal of Automation and Computing | Year: 2015

Cyber-physical systems (CPSs) are integrations of networks, computation and physical processes, where embedded computing devices continually sense, monitor, and control the physical processes through networks. Networked industrial processes combining internet, real-time computer control systems and industrial processes together are typical CPSs. With the increasingly frequent cyber-attack, security issues have gradually become key problems for CPSs. In this paper, a cyber-physical system security protection approach for networked industrial processes, i.e., industrial CPSs, is proposed. In this approach, attacks are handled layer by layer from general information technology (IT) security protection, to active protection, then to intrusion tolerance and physical security protection. The intrusion tolerance implemented in real-time control systems is the most critical layer because the real time control system directly affects the physical layer. This novel intrusion tolerance scheme with a closed loop defense framework takes into account the special requirements of industrial CPSs. To illustrate the effectiveness of the CPS security protection approach, a networked water level control system is described as a case study in the architecture analysis and design language (AADL) environment. Simulation results show that 3 types of injected attacks can be quickly defended by using the proposed protection approach. © 2015, Institute of Automation, Chinese Academy of Sciences and Springer-Verlag Berlin Heidelberg.


Hu X.,Huazhong University of Science and Technology | Hu X.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Fang J.,Huazhong University of Science and Technology | Xiong W.,Huazhong University of Science and Technology
2014 International Conference on Connected Vehicles and Expo, ICCVE 2014 - Proceedings | Year: 2014

A mathematical model is introduced for the IEEE 802.15.4 with unslotted CSMA/CA for mobile vehicle. We use the M/G/l queuing theory to model the mechnism. Our analysis results show the influence of the vehicle speed and the optimum application scope of IEEE 802.15.4 in ITS. © 2014 IEEE.


Xiong W.,Huazhong University of Science and Technology | Hu X.,Huazhong University of Science and Technology | Hu X.,Key Laboratory of Ministry of Education for Image Processing and Intelligent Control | Wang B.,Huazhong University of Science and Technology | Fang J.,Huazhong University of Science and Technology
2014 International Conference on Connected Vehicles and Expo, ICCVE 2014 - Proceedings | Year: 2014

In Intelligent Transport System (ITS) there is a new kind of ad hoc network: Vehicular Ad-hoc Networks (VANET). In VANET, vehicles get useful information by communicating with each other or with a roadside unit. The location information is important in VANET. It is easy to obtain the locations of vehicle by GPS technology. However, there still are some unexpected problems such as not always being in signal coverage or not being strong enough for some applications. It is necessary to develop additional localization techniques to overcome GPS limitations. In this paper, we discuss localization techniques via V2V and V2R communications in VANET when there is no available GPS information, design a dynamic positioning system based on IEEE802.11p and IEEE1609 standards, and simulate positioning algorithms of DV-Hop, RSSI and MDS-MAP. © 2014 IEEE.

Loading Key Laboratory of Ministry of Education for Image Processing and Intelligent Control collaborators
Loading Key Laboratory of Ministry of Education for Image Processing and Intelligent Control collaborators